Liquid crystal displays can be either transmissive or reflective. Their basic construction consists of a liquid crystal material, which is a form of an electro-optic layer, between two substrate plates which have conducting electrodes on their inner surfaces. At least one of the electrodes is a transparent electrode, consisting of a high refractive index material such as indium tin oxide (ITO) placed on top of a transparent substrate such as plastic or glass, with a lower refractive index. Other layers of lower refractive index, such as a passivation layer and a polyimide alignment layer, may be placed on top of the ITO electrode. Because of the refractive index mismatch between these materials, a certain amount of reflection occurs at the interfaces between the liquid crystal and the ITO layer and between the ITO layer and the substrate material. This results in multiple reflections occurring inside the liquid crystal cell which can constructively or destructively interfere depending on the cell gap and the wavelength of the light. This causes highly visible and undesirable colored interference fringes appearing on the display when the cell gap is non-uniform, especially when the cell gap is relatively thin and the illumination spectrum consists of one or more narrow band peaks.
Liquid crystal displays, and in particular liquid crystal on silicon displays, can suffer from problems if the liquid crystal cell gap is made non-uniform during construction. Alternatively, the gap may be constructed uniformly but can subsequently be subjected to stress that can cause a distortion such that the cell gap is not uniform over the area of the liquid crystal display. Typically the liquid crystal layer is only a few microns thick, which is a distance scale that can result in optical interference patterns being formed with many light sources, including LED illumination. In a reflective liquid crystal cell, where the effect of non-uniformity is doubled, a change of the order of 0.2 microns is enough to cause an interference fringe. Indeed, this problem could be more serious than any other visible effect of the underlying cell gap non-uniformity, and so this phenomenon can result in a high reject rate. The fringes can be eliminated by making the cell gap extremely uniform, but this is difficult to achieve with a high yield using present day manufacturing techniques.
Fringes occur because optical interference inside the cell (sometimes enhanced by a polarization effect) changes the amount of light reflected from the display. This interference is a function of cell gap, so changes in cell gap (that would not otherwise be enough to cause problems in other ways) show up as changes in brightness. In certain displays, this change will be much worse for certain colors and so fringes may only be seen in images having those colors, such as red images. FIG. 1 is an illustration of this effect. The peaks on the curve 12 are separated by about 0.2 microns in cell gap, and the underlying intensity change as a function of thickness is small enough that a smooth variation of that amount would not typically be a problem. The graph 10 shows the red intensity as a function of the cell gap. Dot 14 on the curve 12 represents the red intensity for a particular pixel and the dot 16 represents the red intensity of a nearby pixel which would otherwise display the same red intensity as the pixel represented by the dot 14 except that the cell gap for this pixel differs from the cell gap of the pixel represented by dot 14. If these pixels are reasonably close together, then this is typically seen as an objectionable fringe. The fringes give a contour map of the cell gap, with the transition of light to dark representing 0.1 microns and a full fringe dark-light-dark representing 0.2 microns. Clearly it would be advantageous if these variations of the cell gap did not cause such visual artifacts.
Interference fringes, in general, are reduced by suppressing at least one of the reflections that are required to form two interfering beams. In a reflective display, the only component that can be suppressed is the reflective beam component from inside of the glass cover, where the transparent conductive electrode is located. Multi-layer coating techniques provide one way to reduce interference fringes. U.S. Pat. No. 5,570,213 describes a way to reduce the interference fringes by adding additional layers on either side of the ITO layer. These additional layers act as a broadband antireflection coating which effectively refractive index matches the ITO layer to the substrate material on one side and the liquid crystal on the other side. While these layers will decrease the intensity of the observed interference fringes, they are not completely satisfactory because, being a birefringement material, the liquid crystal has two principal refractive index values and it is not possible to simultaneously index match to both of these indices over sufficiently broad spectral range. Furthermore, these antireflection coatings can contain up to 20 different dielectric layers which can be quite expensive to manufacture.
A different approach to eliminate the colored fringes caused by cell gap variations is taken by U.S. Pat. No. 4,693,559. In this case the substrate is roughened with a plurality of depressions, prepared by etching or embossing. The thickness variation within each depression produces a color variation of substantially the entire color spectrum which the eye averages out to a neutral additive color mix since the depressions are relatively small in size. While this method is effective at eliminating fringes, it does introduce a considerable amount of light scattering due to the roughened surface. This roughened surface is the ITO layer. This would make this method unsatisfactory to use in optical configurations where light loss due to scattering cannot be tolerated, such as in projection applications.
A related approach cited in U.S. Pat. No. 5,418,635 adds a plurality of convex portions of two or more different heights formed from photoresist bumps and then covers them with a polymer resin film to give the surface a continuous wave shape without any flat portions. Because there are no flat portions between the top and bottom of the liquid crystal layer, the multiple reflections causing the interference colors cannot occur. While this method has been demonstrated to be effective in reducing interference colors, it suffers from the same scattering limitation of the previous example with the roughened surface.
U.S. Pat. No. 4,632,514 provides a different cell gap under the red, green and blue color filters which is proportional to the dominant wavelength of each of the filters. Thus each separate pixel for these three colors has a different cell gap. In one example from this patent, a 5.4 micron gap is provided under the red filter, a 4.8 micron gap under the green filter, and a 4.0 micron gap under the blue filter. Multiple cell gap color displays provide an improved contrast and viewing angle compared with color displays only having a single cell gap. In this prior patent, the cell gaps themselves are designed to be proportional to the dominant wavelength of each of the filters and thus this design is limited to use in a color display and would not be useful in color displays where colors are generated by other methods, such as time sequential color.